The Failure at the Seams

The Industry Builds Parts; People Inhabit Wholes

The reason is not incompetence. It is structure, almost organizational in nature. The business of making environments is divided into components and the specialists who own them. The audiovisual integrator owns the displays and the sound. The IT department owns the network. A controls vendor owns the building systems; a furniture dealer owns the workplace; a security contractor owns access and cameras; and a facilities team inherits whatever is left when the others have gone home. Each is an expert in a slice. Each is held responsible for the performance of that slice and judged on it. No one's job, on the entire roster, is the whole.

You can watch the gap open in the handoffs. Each contractor arrives with a defined scope of work, executes it, tests the part they were hired to deliver, secures a sign-off, and leaves. The integrator proves the display works. The network team proves the network works. Every signature is honest, and each one covers only the slice it names. The seams between the slices—where a visitor's laptop is supposed to meet the room's display across a network the room does not control—fall outside the scope of work and do not appear on any punch list. They get discovered later by the person standing at the front of a meeting, which is the worst possible place and time to be discovered.

And yet the person who walks into the room does not experience a slice. They do not encounter the audiovisual system, then the network, then the controls, as separate things to be evaluated on their own merits. They encounter one place, all at once, and they judge it as one thing: did this work, did it help me, would I choose to be here again? The org chart is organized by component. Human experience is organized by the whole. The distance between those two ways of organizing the world is precisely where environments fail—not in the parts, which are someone's responsibility, but in the whole, which is no one's. An environment assembled from excellent components by people who were never asked to think about the result will work on paper and, in person, reliably and forever disappoint.

This is also why spending more rarely fixes it. An organization unhappy with its environment that responds by buying more capability—newer displays, another platform, a smarter sensor—usually ends up with more parts to fall out of step with one another, not a better whole. The trouble was never a shortage of components. It was the absence of anyone designing the experience for those components, and no purchase order can make that happen.

We Have Made This Mistake Before

This is not the first time the field has mistaken a thing for its parts. Reyner Banham, writing in 1969, leveled a strikingly similar charge at his own discipline: architecture had celebrated structure and facades, the visible art of enclosure—while very nearly ignoring the mechanical systems that make a building fit to occupy: the heating, the ventilation, the lighting, the controlled provision of comfort that turns a shelter into a place a person can work or live in. The well-tempered environment, in his phrase, was produced at least as much by its services as by its structure, and yet those services were treated as beneath the architect's notice, the unglamorous plumbing of some lesser expertise.

Banham's argument was that the building had always been more than its architecture; it was the environment that architecture and its hidden systems together produced. More than half a century later, we are making the same error, simply moved up one level. We now treat the environment as its visible technology—the screens, the networks, the sensors, the systems marketed as smart—and we ignore the thing all that is supposed to produce: the experience of the people inside. The plumbing has changed. The mistake has not. We keep confusing the apparatus with the result.

The Machine That Taught Us to Think in Parts

Some of where this comes from is an idea so successful that it became invisible long ago. In 1923, Le Corbusier handed the century its governing metaphor: a house, he wrote, is a machine for living in. In its moment, it was liberating. It swept away ornament and sentiment, told the discipline that its history of walls and columns was not the whole story, and insisted that a building be engineered to perform its function as cleanly and efficiently as a well-made machine—a radical, clarifying demand. But a metaphor does more than describe; it teaches a discipline what to value. The machine metaphor taught generations to prize function, efficiency, optimization, and performance, and to see a building as an assembly of systems, each one to be made to do its job as well as possible.

That is, almost exactly, the habit of mind that hands us superbly engineered components and no one attending to the whole they form. A machine is the sum of its optimized parts; tune each component, and the machine runs better. An environment, it turns out, does not behave that way. The thing we are trying to make is not a more efficient machine. It is a better human experience, and experience stubbornly refuses to resolve into the sum of well-tuned parts. A room in which every system is individually optimal can still be a room no one wants to be in. The machine metaphor, for all it gave us, cannot see that because, by its own logic, it should be impossible.

"Smart" Was Always a Word About the Systems

Which brings us to the word that has done more than any other to muddle the matter: smart. The vocabulary of the smart, or intelligent, building is far older than it sounds, and its origins are revealing. The term "intelligent building" was coined in 1981 by a corporation that sells building systems, and two years later, a tower in Hartford, Connecticut, was introduced as the first of its kind. What earned it the name was the integration of its services—climate, lighting, security, fire safety, telecommunications—onto a shared network, and their automation from a central point. The earliest formal definitions of the intelligent building, tellingly, described it wholly in terms of its automated systems and made no mention whatsoever of the people who would spend their days inside it. From birth, smart was a word about the building's systems rather than the occupants' experience—and, just as plainly, a word coined to sell. In the decades since, it has been stretched across so many different things that it now certifies almost nothing: a building is called smart because it has an app, or a dashboard, or a sensor, with no agreement anywhere on what the intelligence is for.

There is a deeper trouble with the word, though, and it is worth stating precisely, because the distinction underneath it runs through everything that follows. What the smart building offered was automation: systems that carry out predetermined rules. If the clock says eight, raise the temperature; if the sensor detects no motion, turn off the lights. Automation is not intelligence. Norbert Wiener, who founded the science of these things, defined an intelligent system by feedback—one that senses its surroundings, compares what it finds to a goal, adjusts its behavior to close the gap, and keeps adjusting as conditions change. A thermostat slaved to a schedule is automated. A system that learns how a space is actually used, recognizes when its own assumptions are wrong, and adapts toward a human outcome is categorically different. Most of what has been sold as smart for four decades is the first thing wearing the name of the second. The difference between automation and real, feedback-driven intelligence is the difference between a building that merely runs and an environment that responds, and the chapters ahead will return to it again.

A Different Place to Begin

So this book begins somewhere other than the component or the pile of components. It begins with the environment as an integrated, experiential whole—the place as the person inside it meets it, all at once, to be judged by what it lets them do, and feel, and become. This is not a dismissal of the parts. The audiovisual systems, networks, controls, and building services are real, demanding, and hard to do well, and the chapters ahead take them seriously. It dismisses the idea that the parts are the point. They are the means. The experience is the end.

It is worth being clear-eyed about why naming the end matters, and Marshall McLuhan is a useful guide here. His enduring insight was that a medium is never neutral—that its form shapes what happens within it quite apart from any content it carries. An environment is a medium in exactly this sense. It shapes the work, the care, the learning, the gathering that take place inside it, whether anyone intended it to or designed it to. The shaping happens regardless. The only real question is whether it is surrendered to accident—the accumulated residue of components installed by people minding their slices—or taken up, on purpose, as the work itself.

What the Whole Is For

To take it up as work, we must be willing to say plainly what the environment is for. That turns out to be harder than it sounds because it is not a technical question, and this industry's reflexes are overwhelmingly technical. It is far easier to specify a system than to state a purpose. But every decision that follows the purpose depends on it—every system chosen, every dollar committed, every standard adopted—and an environment built without an answer is only ever a collection of working parts arranged in a room, waiting to disappoint someone who walks in with a need no one thought to consider.

Before we can usefully talk about how to build environments that succeed as wholes rather than as parts, we have to talk about what the whole is in service of. The answer is human experience—and the argument that experience is not a soft amenity layered on top of the real work but is itself the infrastructure, the very thing all the rest exists to produce, is where we turn next.